Materials and methods

Distribution, description and biology of the species studied. Furcraea andina and Agave americana are the only species of Agaveae that grow wild in the Ecuadorian Andes, where they coexist (^{Jørgensen & León-Yánez 1999}). A. americana is a native of Mexico and the southern United States and considered introduced in Central and South America; it can be found in all of the Andean countries from Colombia and Venezuela to Bolivia (^{eMonocot 2017}, ^{GBIF 2017}, ^{Tropicos 2017}). The distribution of F. andina is restricted to Ecuador, Peru and Bolivia (^{eMonocot 2017}, ^{GBIF 2017}, ^{Tropicos 2017}). Both A. americana and F. andina are long lived, monocarpic succulent plants that form rosettes. Their rosettes grow for several years, and after developing an immense inflorescence, they die (^{Rocha et al. 2006}). These species are also capable of clonal reproduction with colonies surviving for several generations (^{Rocha et al. 2006}).

Vegetative propagation of A. americana occurs mainly through offshoots that emerge from its rhizome system (^{Gentry 1982}). The plant is a wide rosette of tough leaves that reach 5 m in diameter and 2 m in height; leaves have serrated edges, and straight, curved (or slightly curved) thorns, and a large, sharp apical thorn; it takes between 7-17 years for the rosette to reach its reproductive phase (^{Cerón 1994}, ^{Pardo-Briceño 2005}). The time of blooming, and subsequent sap harvesting, can change depending on its care: water and nutrients speed up the process while pruning the leaves delays it (de la Torre et al. in prep). The inflorescence is a panicle that can reach heights of up to 10 meters, with dense clusters of rigid green-yellow flowers (^{Cerón 1994}: Figure 1). In Ecuador, these are pollinated by hummingbirds (Chaetocercus mulsant, Chlorostilbon melanorrhynchus, Colibri coruscans, Lesbia victoriae and Patagona gigas), as well as insects, including bees, bumble bees, butterflies, and dragonflies (^{Cerón 1994}, ^{Ortiz 2003}). Birds, such as Parabuteo unicinctus and Falco sparverius, perch on the flower stalks (pers. obs. of L. de la Torre).

Figure 1 Agave americana and Furcraea andina in the Ecuadorian Andes. A. Flowering rosettes of Agave americana in Andean dry vegetation. B. Flowers of A. americana. C. Flowering rosettes of Furcraea andina in Andean dry vegetation. D. Flowers of F. andina. 

Vegetative propagation of F. andina occurs through bulbils (^{Bailey 1915}). The plant is a green rosette whose leaves while weaker than those of A. americana, are similarly sized, and have spines along the edge of the entire leaf, or sometimes only from the middle section to the tip; leaves have a hard and pointed apex (^{Bailey 1915}, ^{Cerón 1994}). This species generally produces its inflorescence in its third year but, reproduction can be delayed by many years depending on care (^{Bailey 1915}, ^{Cerón 1994}). The plants from which leaves are harvested to obtain fibers can last decades without flowering (R. Rodríguez 2016 pers. comm.). The inflorescence has a panicle-like appearance, but is actually a raceme, typically 10 to 12 m in height. The flowers are aromatic, bell-shaped pendulums that are greenish or yellowish white (^{Cerón 1994}; Figure 1). It is hypothesized that the flowers of Furcraea species are pollinated by insects (^{Rocha et al. 2006}), however, little is known about this due to the low rates of flower visits (^{Eguiarte et al. 2000}). Hummingbirds such as Colibri coruscans and Patagona gigas have been observed to feed on the flowers in the morning in the Ecuadorian Andes (S. de la Torre pers. comm.).

Ethnobotany and productive initiatives. Information was compiled from two sources: 1. Ethnobotanical literature and herbaria specimens documented in the Encyclopedia of Useful Plants of Ecuador (^{de la Torre et al. 2008}). The encyclopedia draws from a collation of ethnobotanical research that has been carried out in Ecuador since the 18th century in at least 100 communities and 14 ethnic groups, including the Kichwa and mestizo communities of the Andes (^{de la Torre et al. 2008}). This body of knowledge has been updated with the most recent publications and unpublished documents containing information on the uses of Andean Agaveae. 2. Semi-structured interviews to stakeholders of productive chains of the species studied. Interviews were conducted in June and July of 2016 across the Ecuadorian Andes (Table 1) and included a common set of open-ended ethnobotanical questions. A network of informants was created starting with representatives from governmental agencies, non-governmental organizations, and the academic sector involved in the promotion of Agaveae-based initiatives, who were asked about productive initiatives they were aware of. These informants identified centers of production and use of Agaveae in the northern, central and southern sub-regions of the Ecuadorian Andes, and constituted 19 % of the total number of interviews (n = 37). Eighty one percent of interviewees were producers from these production centers. Of those interviewed, 65 % were men and 37 % were women (n = 37); 57 % were indigenous people (Kichwa of the Andes), and 43 % were mestizos.

A database of the uses of each species was created. For A. americana we compiled two datasets: 1) uses of the plant, and; 2) uses of the sap, known as mishki. This distinction was made due to the diversity of mishki-specific applications. Each record of use in the database consisted of: species, part of the plant used, description of use, location, and source of information. The uses were classified into nine categories in accordance with ^{Cook (1995)}, with modifications (Appendix 1). A distinction was made by the interviewees in their answers between past uses (that are no longer practiced; these uses were reported by interviewees who recalled that they were practiced in the past), current uses (those still in practice), and potential uses (possible applications that are not currently practiced or are beginning to be tested).

We measured the following parameters of the two species studied: 1. Use diversity, consisting of the total number of uses (past, present, and potential) compiled for each one. 2. Use frequency, equal to the number of use records for each species in the database, that is, the total instances recorded throughout the study for each species. 3. Variation in their use patterns throughout the Ecuadorian Andes (use pattern understood as the distribution of use records in categories). This was analyzed with a chi square test. For the analysis, the Andean provinces were grouped into northern, central, and southern sub-regions; representing the three centers of use of Agaveae reported in the interviews. Possible bias due to the differences in the number of interviews performed in each sub-region was reduced as the database and analyses also included information from literature and herbarium specimens gathered throughout the history of ethnobotany research in Ecuador. Throughout the interviews, Agaveae-related productive initiatives in Ecuador were identified and located. Information about production sites, product processing, and access to the natural resource or raw material was compiled.

Results

Use diversity and use frequency were higher for A. americana than for F. andina. Fifty-two uses were recorded for the different parts of the A. americana plant plus 72 uses for the sap or mishki for a grand total of 124 uses of this species in 548 use records. Thirty-six distinct uses in 140 use records were recorded for F. andina (Tables 2 and 3, Figure 2). All of the Kichwa groups inhabiting the Andean provinces utilize and grow the species studied. The use of F. andina is less diverse and widespread in the southern Andes than in the northern and central Andes; however, we found no significant differences in the number of records by use category between these sub-regions. While no significant differences as regards to the plant of A. americana were found either, we did find differences in the use patterns of mishki. Significant differences were found in the number of records per use category between the northern, central, and southern Andean sub-regions (χ210 = 23.98 p = 0.0076). Previously undocumented common names for both species were recorded (Table 4).

Figure 2 Comparison of the uses of the parts of the plant of A. americana and F. andina. For a description of uses see table 3. *Soap and shampoo. **Medicinal for organic and spiritual diseases. Elaborated by: Nicolás Santibañez and Lucía de la Torre. 

Ethnobotany of Furcraea andina. Its uses were included into eight categories (Table 3). Materials was the use category with the highest diversity of uses and number of use records. Within this category, the most common use is that of its leaf fibers in the production of handicrafts (e.g., bags, hats and footwear. Figure 3) and ropes used in house construction, to tie up livestock or tie crops (27 % of records, n = 140). The practice of cultivating this species as a live fence or boundary marking is found across the region studied, although, according to interviewees, to a lesser extent than in the past and than A. americana. This use does not preclude the plant from being utilized for other purposes like the harvest of its leaves for fibers.

Figure 3 Espadrilles made with the fiber of Furcraea andina. 

More of the uses of F. andina were reported as “past” (28 %, n = 36) than “current” (26 %) (Table 2). According to the users interviewed, the fibers from the leaves are used less and less to make handicrafts and tie up crops such as tomato and cucumber. The fiber’s use to tie up mattresses stuffed with kapok (Ceiba pentandra) cotton dates back more than 50 years. Likewise, the plant’s use as shampoo, laundry soap, and as a treatment for scabies is only remembered but no longer used by the interviewees. Seven potential uses were recorded. These included new applications of the fiber, such as the production of paper (e.g., money, absorbent and artisanal paper), boards or linings, as a substitute for fiberglass, or the use of leaves to make fertilizer or a balanced feed for livestock. There is currently no initiative planning to test these potential uses in the near future.

Ethnobotany of Agave americana. Plant uses were included into nine categories (Table 3). The category with more diversity of uses and records was Materials. Representative uses include: the plant heart in handicrafts, such as chairs or drums (6 % of records, n = 291), the floral stalk in construction and the leaf fibers for handicrafts, such as bags and hats, ropes to tie up livestock, roofs and mudbrick walls made out of sigse (Cortadeira nitida) and floral stalks (5 % of records for each one). Houses with this type of walls were built 50 years ago in the highlands of the Northern Andes and would last up to 15 years. Agave americana is cultivated throughout this region as a live fence to delimit land for grazing and farming (Figure 4). This is the use of the plant with most records (30 %, n = 291), this practice remains current and, like F. andina, does not exclude its other uses. The use of the leaves as livestock feed is one of the primary feedstocks in the central Andes. One agave plant can provide weekly fodder to up to four cows. The most common use as food of A. americana, other than mishki, are its flower buds, known as capers. This use was also reported for F. andina. While not related to the traditional caper, Capparis spinosa, used in Mediterranean cuisine, the Agaveae caper has a similar appearance and flavor after it is pickled in vinegar, lemon or chicha. Capers are a traditional specialty during Carnival in the city of Guaranda, and thus included in the Social use category as well.

Figure 4 Live fence of Agave americana delimiting crop lands in Salasaka (Central Andes of Ecuador). 

We recorded 17 past uses for the plant (33 % of the total uses, n = 52) while 21 uses remain current (40 %; Table 2). The leaf fibers and plant parts that were used as shampoo, soap, dye, toys or blackboards have been almost totally replaced by synthetic products. Seven potential uses are part of the initiatives that are starting to emerge. These include the use of the leaves to produce flour for human consumption, compost and food for livestock, canned flower buds, shampoo made with the root at an industrial level, the whole plant for soil restoration in agroforestry systems and as biofuel.

Ethnobotany of mishki. Mishki means “sweet” in Kichwa and, as noted, is the name given to the sap that is obtained by exudation from the heart of the agave plant. It is obtained through common practices throughout the Ecuadorian Andes (Figure 5). Producers identify the maturity of the agave plant; as it gets larger, the heart thickens relative to the rest of the plant, the leaves broaden, lie flat, snap when crushed with the fingers, become whitish and their base yellowish, and the shoot shrinks and becomes thinner. Three smaller, narrower leaves with a curved apical spine grow around the shoot. The spine is known as pico de loro (parrot’s beak) or pico de gorrión (sparrow’s beak). July to September is the preferred time for harvest given the lack of rain that would otherwise damage the mishki. This is also when mishki is most in-demand owing to a lack of water from natural sources. Users recognize and cultivate variants of A. americana that are more suitable for harvesting sap, such as those with a higher concentration of sugar and smaller or no thorns on the leaves.

Figure 5 Method of “cutting” the agave's heart by a lateral incision, and a rock used to block the hole at the agave’s heart. These practices are characteristic of the Ecuadorian Andes. 

Uses were included in six categories (Table 3). The use category with the highest number of uses and records is Medicinal (49 % of records, n = 257 and 35 distinct applications). In the Northern Andes, 28 medicinal uses were reported and 15 are exclusive to this region. These vary from the treatment of acne to cancer. In the South, mishki is associated with the treatment of kidney diseases. Consumption to treat conditions of the musculoskeletal, respiratory and digestive systems is practiced throughout the whole country. In some places, the same medicinal properties are reported for the syrup and liquor made from this exudate.

The most common edible uses of mishki are as a sweetener, thirst quencher, and as an invigorating beverage. Among the indigenous communities in the Andes, mishki was the only sweetener available before the arrival and cultivation of sugar cane (Saccharum officinale). Traditional beverages such as coladas (thick beverages) from varieties of corn, Cucurbita maxima, Curcubita pepo, Oxalis tuberosa, were prepared with this sweetener. New proposals for mishki and its syrup include gourmet products like sauces and marmalades.

The Social use category includes the consumption of guarango (fermented mishki), a liquor (distilled guarango), and to a lesser extent, drinks made of guarango and fruit, known as curado or pingolito, during parties and mingas (communitary work). The production and consumption of guarango and liquor is higher among the Cayambis in the north and Cañaris and Saraguros in the south.

Although the use of mishki as Animal food, specifically for pigs, is present in the entire Andes, it is more common in the south. In some areas this is the only application of mishki. Most of the uses for mishki are still current. Six were recorded as “past” uses, which are applications that have disappeared in some areas but remain alive in others (e.g., nutritious food for women after childbirth, for longevity and vitality for aged persons and as a substitute for water in arid regions). The only use that has disappeared throughout the entire country is the consumption of the dried and toasted bran that is obtained by scraping the orifice of the plant heart to obtain the mishki. This food was considered to be extremely nutritious and was consumed as an alternative when the crops were damaged due to frost in the province of Cotopaxi. Thus, it was vital for the survival of the people of the region. Only one potential use of mishki was documented: bottled and pasteurized mishki. It is expected to be sold at a national level and this initiative has already begun to be implemented in Salasaka.

Productive Initiatives. Fifty-three productive initiatives were identified with Agaveae in the Ecuadorian Andes (Figure 6). More startups have been developed in communities in the northern (26) and southern (17) sub-regions rather than in the central (10) sub-regions. Most of them (92 %) include A. americana, and mishki. Thirteen products are commercialized in Ecuador. Those most frequently marketed are mishki and its by-products (e.g., barley rice with mishki, syrup and liquor). The production of fiber and handicrafts is considerably lower and involves fewer actors (Table 5). Nine producers considered that there is a scarcity of A. americana plants due to: 1- low rates of replanting of harvested plants; 2- the construction of roads that has destroyed this plant’s habitats; and, 3- the loss of traditional knowledge and uses of the species. Last year, the production of packed capers decreased six fold during the carnival in Guaranda due to scarcity. This has become more noticeable in the last years in areas where liquor is produced in the Southern Andes.

Figure 6 Location of productive initiatives with Agaveae in Ecuador. 1: Northern, 2: central and 3: southern sub-regions of the Andes. 

Discussion

Cultural importance of Andean Agaveae. Agave americana and Furcraea andina can be considered keystone species for the development of Andean indigenous and mestizo cultures of Ecuador. These species played an important role as a source of materials in the past. A. americana played and continues to play a fundamental role in people’s diets, livestock breeding, as a medicine and in ceremonies. For example, mishki was consumed as a substitute for water to quench thirst during dry seasons or in dry areas, it was the main sweetener before the arrival of sugar cane and the bran obtained from scraping the heart was consumed as an alternative food source when frost destroyed crops. The fermented mishki or guarango is an indispensable drink during celebrations such as the Inti Raymi (grain harvesting season during the summer solstice) for Kichwa groups. The flower buds or capers are consumed as a traditional specialty during Carnival in the city of Guaranda, and are also consumed during Easter in rural communities (Cobo 2014). This species promotes cultural cohesion in the Ecuadorian Andes.

Agave americana higher use diversity and versatility lies in its mishki-producing heart. In turn, usefulness of more versatile species is generally more widespread among people than that of less versatile ones (^{Tardío & Pardo-de-Santayana 2008}, ^{Macía et al. 2011}). Furthermore, A. americana is one of the plants most cultivated as a live fence in the Andes (^{Carlson & Añazco 1990}), which would contribute to its high number of medicinal uses that it has. Due to its availability, there would be more opportunities to experiment with new medicinal applications for this species (^{Stepp & Moerman 2001}, ^{Bjorå et al. 2015}). F. andina is also a versatile plant, although to a lesser extent because its “heart” is considered too narrow to obtain mishki. It is cultivated less (de la Torre et al. in prep.) and, therefore, its availability is reduced. Agave americana meets all the criteria to be considered a cultural keystone species given its high rating according to an index based on six criteria proposed by ^{Garibaldi & Turner (2004)} and outlined

below:

Furcraea andina was fundamental as a source of fibers used in hunting, cultivation, clothing, footwear, household goods and construction. It could be considered to have been a keystone species, particularly in the past before there were synthetic materials available to replace its fibers. The fibers of A. americana were used for similar purposes but F. andina fibers were preferred as they are easier to obtain from the leaf, to dye, and to weave owing to their softer and more flexible nature, furthermore, they can be harvested in less time (M. Presentación and Z. Guamán pers. comm.). The dependency of cultural keystone species on context (temporal, spatial and social) has been addressed by the creators of this concept (^{Cristancho & Vining 2004}, ^{Garibaldi & Turner 2004}). F andina, could have played a key role within a restricted time period in the past for Andean cultures in Ecuador.

The reported cultural value of A. americana and F. andina can contribute to the conservation of its associated culture and could enhance initiatives for Andean ecosystem conservation and restoration with these species.

Agave americana, an introduced species in Ecuador? The cultural importance of A. americana throughout the Ecuadorian Andes supports the hypotheses that it arrived long ago. The significant differences found in the use patterns of mishki between the northern, central and southern sub-regions of the Ecuadorian Andes, in addition to evincing the versatility of A. americana, could indicate that it has been subject to an extended period of use. Plant use patterns in a given place depend on interrelated natural, social, cultural and historical factors. These factors include species abundance, varietal characteristics, plant diversity of the surrounding ecosystem (allowing access to substitute species), cultural preferences, the technology available to process that resource, market access, time of arrival, i.e., time for experimentation with the plant, and contact with other human groups resulting in the transfer and loss of knowledge (^{Campos & Ehringhaus 2003}, ^{Byg & Balslev 2004}, ^{Byg et al. 2007}). Thus the variation in use patterns found imply the extended interaction of these factors. The archeological evidence of vessels suggesting their use for storing mishki and its by-products (^{Proyecto Chasqui 2016}, Mindalae Museum pers. obs. of L. de la Torre)

and testimonials of chroniclers (^{De Velasco 1977}, ^{De la Vega 2004}) also support this hypothesis.

Despite its continuous distribution in America (^{eMonocot 2017}, ^{GBIF 2017}, ^{Tropicos 2017}), and its proven capacity for natural propagation (^{BioNET-EAFRINET 2001}, ^{Hodgkiss 2017}, A. americana is considered an introduced species in South America (^{Peña 1976}, ^{Tirado 2002}, ^{Venero-González 2006}, ^{Bernal et al. 2015}), although there are no studies specifying how and when it was introduced. There are native species of Agave in other Andean countries: Agave cocui in Venezuela and Colombia, A. cudinamarcensis, A. boldinghiana, A. pax and A. wallisii in Colombia and A. cordillerensis in Peru (^{Lodé & Pino 2008}, ^{Hoschstatter 2015}, ^{Hodgkiss 2017}, ^{Giraldo-Cañas 2017}). It is plausible that other native species of Agave or at least native varieties of A. americana exist in Ecuador. It would be desirable to perform genetic analyses that clarify its arrival to South America, and, particularly, to Ecuador.

Besides the contribution to the knowledge of the origin and distribution of agaves that these investigations would represent, A. americana would benefit from political decisions that prioritize native taxa over introduced ones, e.g. reforestation incentives, the declaration of its products and processing cultural patrimony, and the promotion of productive initiatives.

Use of Agave in the equatorial Andes compared to Mexico and other Andean countries. The cultural importance of A. americana in the equatorial Andes is comparable to that found in Agave spp. in Mexico, but not to other Andean countries where its uses are less diverse, less extended and sporadic (see Tables 6 and 7 for citations and a comparison of uses). In Mexico, there are 186 taxa within the Agave genus that grow in 75 % of the territory; more than 100 species are useful and have played a fundamental part of the survival of communities in arid and semiarid areas (^{García-Mendoza 2007}, ^{García-Herrera et al. 2010}, ^{Colunga-GarcíaMarín et al. 2017}). At least 12 species are domesticated (^{Colunga-GarcíaMarín et al. 2017}). Terms for these species and for each part of the plant can only be found in some languages in Mexico and the Kichwa of the Ecuadorian Andes (Table 4, Appendix 2).

There are similarities and differences in the uses given in Ecuador and Mexico (Table 7), likewise in the mishki-obtaining procedures and by-product preparation (Appendix 3). It is possible that there was an exchange of knowledge between Mexico and Ecuador. This could have occurred either if the plant had arrived by its own means or was introduced by humans due to its utility. This exchange could have happened at a number of different times:

The comparison of Agave uses between Ecuador and México, highlights the high use diversity and cultural importance of the genus in Ecuador, even though that there is only one species available. It has been reported that in areas of decreased species diversity, a similar variety of uses is distributed among the fewer species available, resulting in higher use intensity per species (^{FAO 1995b}).

Future perspectives of the Andean Agaveae. Furcraea andina has traditionally been a primary source of fiber for indigenous communities in Ecuador, however this use is rapidly disappearing due to the increased availability of lower priced, synthetic products. Market access is one of the main factors that affect plant use patterns resulting in a decrease in the use of vegetable products by local communities (^{Ladio & Lozada 2001}, ^{Byg et al. 2007}).

In Ecuador, the area used for cultivation for Agaveae-fiber production has decreased from around 3,000 hectares in 1980 (^{FAO 1995a}) to around 1,000 hectares in 2013 (^{Ayora & Quito 2013}). As recently as 2008, Ecuador exported 17.39 T of rope and riggings made from Agaveae fiber (^{Yépez et al. 2009}), while in 2013, Ecuador went from exporting Agaveae fiber to importing it (T. Zamora pers. comm.). Fiber production from F. andina and other fiber by-products is limited to a few initiatives, generally organized groups of indigenous women that produce handicrafts of unique beauty and quality but with limited markets. Furthermore, currently no initiative is being tested for the potential applications of F. andina for fiber (e.g., production of artisanal, absorbent or money paper). These uses could benefit the country as it is a net importer of fiber (^{Yépez et al. 2009}, ^{Viveros 2017}).

The number of initiatives with A. americana has increased in recent years due to a number of reasons. These include: 1- the growth of organizations that have supported the development of productive alternatives in rural indigenous communities; 2- the increase in consumption of mishki and its by-products owing to their nutritional and medicinal qualities emphasized by the media (^{AGN 2016}, ^{Agrovida 2016}, ^{Tungurahua 2016}); the production and consumption of liquor has also increased as a result of tasting events being held (Alvarado 2017); and 3- an increasing tendency in consumption of natural products, especially sweeteners that are healthier than those obtained from sugar cane (^{Tandel 2011}, ^{Mellado-Mojica & López 2015}). Coincidentally in areas with more mishki-based initiatives, in the Northern and Southern Andes, there is a major perception of scarcity of A. americana. To avoid overexploitation and local extinctions, it is necessary to replace the harvested plants and to reforest marginally arid areas with this species. As 49,000 km2 or up to 17 % of Ecuador’s surface is suitable for its cultivation, the prospects for increased production of A. americana are positive (de la Torre et al. in prep).

To satisfy the increasing demand of tequila and mezcal, natural populations of Mexican agaves have been overexploited (^{Colunga-GarcíaMarín et al. 2017}), which together with the degradation of natural habitats, has threatened 48 species of agaves (^{Franco-Martínez 1995}). In 2016, more than 115,000 hectares of Agave tequilana were cultivated throughout Mexico (^{SAGARPA 2016}) and 273.3 million liters of tequila were produced (^{Consejo Regulador del Tequila 2017}). The monocrop of clones of Agave tequilana has caused a reduction in genetic diversity and contributed to the destruction of natural habitats (^{Colunga-GarcíaMarín et al. 2017}). Thus, it is recommended that A. americana be cultivated from seeds and not only through offshoots, and in environment friendly systems such as agroforestry systems. Traditional practices of indigenous communities could be taken as a model.

To achieve the sustainability of productive initiatives with A. americana, the associated culture and history behind its by-products must be emphasized. Furthermore, Agaveae products come from plants that grow in arid areas, contribute to soil improvement, prevent erosion, regenerate ecosystems and mitigate climate change through carbon fixation (^{García-Herrera et al. 2010}). Currently 13 products derived from Agaveae are commercialized in Ecuador, few compared to the diversity of uses recorded in this study. Information from this study can encourage the development of new products that help to improve the quality of life of agricultural communities in the Ecuadorian Andes.